resume-tailor

分类: 工具与效率 | 上传者: Zhangyinglun | 下载: 0 | 版本: v1.0（最新）

Use when user provides JD and existing resume, expecting job-targeted optimization, ATS keyword alignment, language polishing, and delivery of a single-page A4 PDF resume.

更新日志: Source: GitHub https://github.com/Zhangyinglun/resume-tailor-skill

目录结构

当前层级: 根目录

📁 .claude/
- 📁 commands/
  - 📄 check-resume-tailor-setup.md 521 B
  - 📄 resume-tailor.md 789 B
📁 .opencode/
- 📁 command/
  - 📄 install-skill-deps.md 758 B
📁 docs/
- 📁 guide/
  - 📄 installation.md 1.3 KB
📁 install/
- 📄 agent-install.yaml 816 B
📁 references/
- 📄 ats-keywords-strategy.md 6.3 KB
- 📄 execution-checklist.md 5.5 KB
- 📄 optimization-actions.md 2.4 KB
- 📄 profile-cache-template.md 1.1 KB
- 📄 prompt-recipes.md 1013 B
- 📄 resume-working-schema.md 1.7 KB
📁 resume_output/
- 📁 backup/
  - 📄 .gitkeep 0 B
- 📄 .gitkeep 0 B
📁 scripts/
- 📄 check_agent_platform_support.py 13.2 KB
- 📄 check_content_quality.py 6.7 KB
- 📄 check_pdf_quality.py 14.1 KB
- 📄 generate_final_resume.py 6.0 KB
- 📄 layout_auto_tuner.py 9.7 KB
- 📄 resume_cache_manager.py 18.1 KB
- 📄 resume_shared.py 7.7 KB
📁 templates/
- 📄 design_philosophy.md 3.2 KB
- 📄 design_tokens.py 1.9 KB
- 📄 layout_settings.py 1.9 KB
- 📄 modern_resume_template.py 14.9 KB
- 📄 README.md 3.6 KB
📁 vendor/
- 📁 skills/
  - 📁 docx/
    
    📁 scripts/
    
    📁 office/
    
    📁 helpers/
    
    📄 __init__.py 0 B
    
    📄 merge_runs.py 5.4 KB
    
    📄 simplify_redlines.py 5.6 KB
    
    📁 validators/
    
    📄 __init__.py 336 B
    
    📄 base.py 31.9 KB
    
    📄 docx.py 16.0 KB
    
    📄 pptx.py 9.6 KB
    
    📄 redlining.py 8.7 KB
    
    📄 pack.py 4.9 KB
    
    📄 soffice.py 5.2 KB
    
    📄 unpack.py 4.0 KB
    
    📄 validate.py 3.6 KB
    
    📁 templates/
    
    📄 comments.xml 2.5 KB
    
    📄 commentsExtended.xml 2.5 KB
    
    📄 commentsExtensible.xml 2.6 KB
    
    📄 commentsIds.xml 2.6 KB
    
    📄 people.xml 115 B
    
    📄 __init__.py 1 B
    
    📄 accept_changes.py 4.0 KB
    
    📄 comment.py 10.4 KB
    
    📄 LICENSE.txt 1.4 KB
    
    📄 SKILL.md 19.6 KB
  - 📁 humanizer/
    
    📄 README.md 7.7 KB
    
    📄 SKILL.md 22.6 KB
    
    📄 WARP.md 2.1 KB
  - 📁 pdf/
    
    📁 scripts/
    
    📄 check_bounding_boxes.py 2.7 KB
    
    📄 check_fillable_fields.py 268 B
    
    📄 convert_pdf_to_images.py 1008 B
    
    📄 create_validation_image.py 1.2 KB
    
    📄 extract_form_field_info.py 4.2 KB
    
    📄 extract_form_structure.py 3.9 KB
    
    📄 fill_fillable_fields.py 3.7 KB
    
    📄 fill_pdf_form_with_annotations.py 3.2 KB
    
    📄 forms.md 11.6 KB
    
    📄 LICENSE.txt 1.4 KB
    
    📄 reference.md 16.3 KB
    
    📄 SKILL.md 7.9 KB
📄 .gitignore 353 B
📄 AGENTS.md 2.5 KB
📄 CLAUDE.md 1.4 KB
📄 LICENSE 1.1 KB
📄 README.md 2.5 KB
📄 README.zh-CN.md 2.5 KB
📄 requirements.txt 36 B
📄 SKILL.md 6.7 KB

SKILL.md

---
name: resume-tailor
description: Use when user provides JD and existing resume, expecting job-targeted optimization, ATS keyword alignment, language polishing, and delivery of a single-page A4 PDF resume.
---

# Resume Tailor Skill

## Core Objectives
- **ATS Hit Rate**: Cover high-frequency JD keywords while maintaining natural expression.
- **Job Relevance**: Prioritize evidence chains and quantified results that best match the position.
- **Delivery Quality**: Output A4 PDF with extractable text and pass format quality checks.

## Inviolable Principles
- **No Fabrication**: Only rewrite, rearrange, compress—no fictional experience/tech/data.
- **Volume Gate Before Generation**: Compress content to single-page target before generating PDF.
- **Autonomous Decision Making**: Agent autonomously makes all optimization decisions (keyword alignment, content prioritization, compression, layout tuning) without asking for confirmation.
- **Transparent Reporting**: All decisions are recorded in the final summary report for user post-hoc review.
- **ATS Friendly**: No table-based main layout, no images replacing body text, no key info in headers/footers.
- **Fixed A4 + Single Page**: Default delivery 210mm x 297mm, 1 page.
- **PDF Only Delivery**: Final output is PDF only.
- **Read `vendor/skills/pdf/SKILL.md` before PDF actions**: Execute before initial generation, minor regeneration, or final export.
- **Auto-fit Scope**: Automatic tuning may adjust layout parameters only; it must not rewrite resume content unless user explicitly asks.

## Stateless Boundary (Mandatory)
- `resume-tailor` skill directory can only store "rules + templates + scripts + references".
- No personalized cache (user preferences, historical JDs, contact info, resume samples) in skill directory.
- Unified directories (relative to workspace):
  - `resume_output/`: Latest PDF
  - `resume_output/backup/`: Historical PDF backups
  - `cache/user-profile.md`: Long-term preference cache
  - `cache/resume-working.json`: Current session resume body
  - `cache/jd-analysis.json`: JD analysis results (keywords, alignment, optimization actions)

## Minimum Execution Flow

### Phase A – Initialize
1. Run `scripts/resume_cache_manager.py reset`.
2. Run `template-check`; if template exists, run `template-use` to load into working cache. If not, run `template-init` with user-provided resume, then `template-use`.

### Phase B – Analyze & Draft
1. **JD Diagnosis**: Analyze JD (or target direction), produce P1/P2/P3 tier classification with gap report, and persist results to `cache/jd-analysis.json` via `jd-save`.
2. **Apply All Modifications**: Apply optimization decisions using standard action codes (`LEAD_WITH`, `EMPHASIZE`, `QUANTIFY`, `DOWNPLAY`, `MERGE`, `REWORD`) as defined in `references/optimization-actions.md`. Document each action with target, code, and reason. Then run `update` to persist.

### Phase C – Compress & Quality
1. **Volume Gate**: Score bullets via `score_all_bullets()` against `cache/jd-analysis.json`, then check working cache against volume thresholds; if compression needed, prioritize removing lowest-scored bullets first, then `update`.
2. **QA & De-AI**: Read `vendor/skills/humanizer/SKILL.md` for natural expression guidance, then run structure/quantification/ATS checks.

### Phase D – Generate & Deliver
1. **PDF Generation**: Read `vendor/skills/pdf/SKILL.md`, then generate with `--auto-fit`. If QC fails, retry up to 3 times with escalating parameters.
2. **Summary Report**: Output a structured summary report covering all decisions made. **Must include the full absolute path of the generated PDF file.**
3. **Wrap-up**: Update `cache/user-profile.md`, retain working cache.

Complete checklist and thresholds in `references/execution-checklist.md`.

## Summary Report Format

After PDF delivery, output the following report:

```
## Resume Tailor Summary Report

### Job Analysis (from cache/jd-analysis.json)
- Position: {title}
- P1 (Critical): {keywords}
- P2 (Important): {keywords}
- P3 (Nice-to-have): {keywords}
- Matched: {aligned keywords}
- Gaps: {gap keywords with strategies}

### Modifications Applied
- Actions applied: {list of action_code → target, e.g., EMPHASIZE → experience[0].bullets[2]}
- Keywords added: {list}
- Content reordered: {yes/no, brief description}
- Content removed: {list of removed items with reasons}

### Volume & Compression
- Before: {word count} words / {line count} lines / {bullet count} bullets
- After: {word count} words / {line count} lines / {bullet count} bullets
- Compression actions: {list}

### QA Results
- Humanizer: {applied / not needed}
- Structure check: {pass/fail + details}
- ATS keyword coverage: {percentage or list}
- PDF QC: {pass/fail + checks summary}

### Deliverable
- PDF path: {full absolute path of generated PDF}

### Action Required
- Review contact info and sensitive data in the generated PDF.
- {Any failed QC checks or caveats, if applicable}
```

## Fixed Template Baseline (Logical Structure)
Always organize content in the following module order:

```markdown
Header
Summary
Professional Experience
Technical Skills
Education
```

## Script Responsibility Boundaries
- `scripts/resume_cache_manager.py`: Manage `cache/resume-working.json` reset/init/update/show (`cleanup` only manual as needed), `cache/base-resume.json` template-init/template-use/template-show/template-check, and `cache/jd-analysis.json` jd-save/jd-show.
- `scripts/generate_final_resume.py`: Accept `--input-json`, optionally `--auto-fit`, and generate final PDF.
- `scripts/check_pdf_quality.py`: Perform general format and text quality checks.
- `scripts/check_content_quality.py`: Content-level quality checks on resume JSON — bullet scoring, verb strength, quantification rate.
- `scripts/resume_shared.py`: Shared utilities (validation, JSON I/O, parsing helpers) consumed by other scripts.
- `templates/modern_resume_template.py`: Only responsible for PDF layout and export.

## Bundled Skills (vendor/)

All dependent skills are bundled in the `vendor/skills/` directory — no separate installation required.

- `vendor/skills/pdf/SKILL.md`: PDF read/generate guide — read before any PDF action
- `vendor/skills/docx/SKILL.md`: DOCX read/edit guide — read before reading `.docx` files
- `vendor/skills/humanizer/SKILL.md`: AI trace removal guide — read before humanizing text

## Reference Materials
- Full process checklist and thresholds: `references/execution-checklist.md`
- ATS strategy: `references/ats-keywords-strategy.md`
- Prompt templates: `references/prompt-recipes.md`
- Profile cache template: `references/profile-cache-template.md`
- Working cache structure: `references/resume-working-schema.md`
- Template instructions: `templates/README.md`
- Optimization action codes: `references/optimization-actions.md`

登录后下载/点赞/收藏 ❤ 2 | ★ 0

resume-tailor

目录结构

SKILL.md

举报内容

提示